Loading articles…
Loading articles…
Blog
Guides from our engineering team on AI agents, workflow automation, custom software and the decisions around them. No hype, just what works.
Featured
Latest
When pgvector is enough for RAG, how filters quietly shrink your results, and the settings, indexes and partitioning that fix it before you add a new database.
How to trace AI agents with OpenTelemetry GenAI conventions: the span tree to emit, attributes worth keeping, what to leave out, and why the spec is still moving.
How to replace a legacy system piece by piece with the strangler fig pattern, and where AI coding agents genuinely help and where they still need a human owner.
Google says AI Overviews need no special markup. The real work is crawler access, snippets and content. What to allow or block in robots.txt, and how to check it.
Million-token windows did not kill retrieval, but they changed the question. How to decide between RAG, long context and prompt caching, with costs and failure modes.
The 2026-07-28 MCP spec removes sessions and the initialize handshake. Here is what breaks, what to change, and the new security duties on your server.
The EU delayed high-risk AI rules to 2027, but chatbot disclosure and AI-content marking rules already apply. A practical checklist for software teams.
What AI agents are, how they differ from automated workflows, where they create value in operations, and how to deploy them with guardrails and oversight.
A practical comparison of n8n, Zapier and Make.com: ease of use, pricing models, hosting, custom code, AI features and when each platform fits best.
When to buy SaaS, when to build custom software, and when a hybrid approach works best, with a practical checklist for comparing cost, fit and risk.
Start a project
Share the workflow, product or team challenge. We will come back with a clear recommendation and next steps.
Intro call · Fixed-scope proposal · No obligation
